»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
Skarsnik m: my uint32 $u32; $u32 = -4; say $u32; 00:00
camelia rakudo-moar 6a8278: OUTPUT«4294967292␤»
Skarsnik m: my uint32 $u32 = 0; $u32 -= 5; say $u32;
camelia rakudo-moar 6a8278: OUTPUT«4294967291␤»
Skarsnik I think all intxx and uintxx are not equal on constraint 00:01
tbrowder p6: my uint64 $n = 232; say sprintf "%0x", $n
camelia rakudo-moar 6a8278: OUTPUT«e8␤»
Skarsnik m: my uint64 $u64 = 0; $u64 -= 5; say $u64;
camelia rakudo-moar 6a8278: OUTPUT«-5␤»
Skarsnik probably a bug
tbrowder say UINT_MAX
p6: say UINT_MAX 00:02
camelia rakudo-moar 6a8278: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared name:␤ UINT_MAX used at line 1␤␤»
tbrowder p6: say UINT32_MAX 00:03
camelia rakudo-moar 6a8278: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared name:␤ UINT32_MAX used at line 1␤␤»
tbrowder p6: my uint32 $n = 0xffffffffffffffff; say $n 00:05
camelia rakudo-moar 6a8278: OUTPUT«4294967295␤»
tbrowder p6: my $uint32-max= 0xffffffffffffffff; my $x = -4; my $xuint = $uint32-max + $x + 1; say sprintf "%x", $xuint 00:09
camelia rakudo-moar 6a8278: OUTPUT«fffffffffffffffc␤»
tbrowder okay, i think i may know how to do it 00:10
00:12 pierre_ joined, MilkmanDan left 00:15 gfldex left 00:19 ItayAlmog left
timotimo oh hey Skarsnik 00:21
00:25 kurahaupo left 00:40 pierre_ left, pierre_ joined, pierre_ left 00:42 MilkmanDan joined 01:00 BenGoldberg joined 01:03 Skarsnik left 01:07 mempko left 01:08 johnjohn101 joined 01:12 MilkmanDan left 01:20 SmokeMachine____ joined 01:23 MilkmanDan joined, mithaldu_ joined 01:24 Lucas_One joined 01:26 BuildTheRobots joined 01:28 pierre_ joined 01:29 kipd joined 01:33 olinkl joined 01:36 sufrostico left, notbenh_ joined 01:37 AngeloMichael joined 01:39 pierre_ left 02:01 dj_goku left 02:12 zacts_pi joined, zacts_pi is now known as zacts 02:14 pierre_ joined 02:45 zacts left, noganex_ joined 02:49 noganex left 02:54 Ben_Goldberg joined 02:55 BenGoldberg left, pierre_ left 03:23 dj_goku joined, dj_goku left, dj_goku joined
skids m: my $f = Channel.new; start { react { whenever $f { .say; $f.close; } }; "done".say }; sleep 1; $f.send(42); sleep 1;' 03:34
camelia rakudo-moar 6a8278: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unable to parse expression in single quotes; couldn't find final "'" ␤at <tmp>:1␤------> 3".say }; sleep 1; $f.send(42); sleep 1;'7⏏5<EOL>␤ expecting any of:␤ single quotes␤ …»
skids m: my $f = Channel.new; start { react { whenever $f { .say; $f.close; } }; "done".say }; sleep 1; $f.send(42); sleep 1;
camelia rakudo-moar 6a8278: OUTPUT«42␤done␤»
skids m: my $f = Channel.new; my $g = Channel.new; start { react { whenever $f { .say; $g.close; }; whenever $g { .say } }; "done".say }; sleep 1; $f.send(42); $g.send(1); sleep 1;
camelia rakudo-moar 6a8278: OUTPUT«42␤1␤»
dj_goku anyone else noticing a lot of fuzzy testing lately? 03:46
I think I have seen 3-4 posts on reddit/hacker news about people fuzzing an app or language. 03:47
timotimo hack.p6c.org/~timo/ :P 03:48
dj_goku lol nice 03:51
timotimo: prior to the recent trend 03:52
timotimo oh, yes
that's true
dj_goku timotimo: is there a reason this isn't done periodically?
btw all I see is numbers not sure what it all means.
timotimo hehe 03:53
yeah, it doesn't mean terribly much
if i really "meant it", i'd write a custom harness that makes it unnecessary to launch a full moar process for each single test case
also, this fuzzing is only for loading (and other fuzzing runs for running) .moarvm files 03:54
with a start-up time of about 0.1 seconds, it's kind of hard to fuzz rakudo with strings for programs
though perhaps i'll give that a try some time :D
dj_goku :D 03:56
04:03 Xliff joined
timotimo here's something mildly interesting 04:04
bind_key is somewhat expensive during startup
(from MVMHash) 04:05
28 strings have more than 100 occurences during startup
04:06 johnjohn101 left
timotimo the "long tail" after that is 5389 different strings 04:06
04:06 bjz_ left
timotimo 1778 of those only occur once 04:06
now, it doesn't seem like we're storing the hash code in a string object after calculating it 04:07
04:08 bjz joined
dj_goku all greek to me. :D 04:08
04:08 Cabanossi left
timotimo oh, huh 04:10
04:10 Cabanossi joined
timotimo seems like perhaps we *are* using the cached hash code ... somehow 04:11
but we're not using the hash code for MVMHash ?!? 04:12
04:23 khw left 04:25 DrForr_ joined
timotimo callgrind seems to think we spend most of our time allocating bits of memory through malloc, then deserialization (first place being serialization_read_int, second place being serialization_read_ref), then validating static frames, then binding data to hash keys, then doing running in the interpreter itself, then allocating stuff in the old generation, then allocating stuff in the garbage collector "in 04:26
general", then more validation of bytecode, then more deserialization
but going by inclusive, the work_loop function (which is what does deserialization) really is a very expensive thing 04:27
even though validate_static_frame is one of the more expensive functions "self"-wise, it's very low in the ranking when it comes to inclusive timings 04:28
04:29 pierre_ joined 04:31 DrForr_ left
MasterDuke timotimo: have you ever re-run your coverage report? 04:33
04:34 bjz left 04:35 pierre_ left 04:39 mempko joined
timotimo not recently. i'm farther past bedtime than i've been in many, many weeks. remind me again after i've had some sleep? :) 04:40
04:41 darutoko joined
MasterDuke will do. and speaking of coverage and fuzzing and such, has anybody seen if coverity or pvs-studio will do a free analysis of MoarVM? they've done a bunch of other open-source stuff projects (e.g., WINE, Libre Office) 04:42
04:45 ufobat joined 04:46 cyphase left 04:51 cyphase joined 04:53 NlkIoUxhGD joined 04:58 setty1 joined 05:13 skids left 05:16 nadim joined 05:28 nadim left 05:30 NlkIoUxhGD left 05:36 firstdayonthejob joined 05:43 uDMCIOLLth joined
timotimo MasterDuke: irclog search reveals that we've had a few changes in moar due to coverity scans already 06:01
06:02 domidumont joined
timotimo moritz: requesting support for "literal" matches; i can't search for "coverity" in #perl6 06:02
moritz: because it finds everything that has "cover" in it, apparently
MasterDuke: github.com/MoarVM/MoarVM/commit/688eecdd22 - example 06:03
06:06 domidumont left 06:07 domidumont joined 06:08 firstdayonthejob left 06:15 mempko left
timotimo .tell [ptc] were you the one who got coverity for moarvm? 06:17
yoleaux timotimo: I'll pass your message to [ptc].
06:18 firstdayonthejob joined
timotimo MasterDuke: scan.coverity.com/projects/paultcochrane-moarvm 06:20
last analyzed was sep 02 2015, but i guess you can just hit the "add me to project" button if you want to put a new scan up, MasterDuke? 06:21
06:24 CIAvash joined 06:28 wamba joined 06:33 rindolf joined 06:37 benjikins joined 06:40 krunen is now known as krunen_ 06:44 nadim joined
ufobat what does the * in '... method new($x){ self.bless(*, :$x)} ...' mean? 06:52
timotimo it means your code is out of date and needs you to remove the "*, " from it :)
ufobat and what did it mean? :) 06:53
benjikins How do I check if a string contains any element in an array
in the shortest possible way
timotimo it meant nothing
ufobat yay :D
timotimo benjikins: you can try $mystring.contains(any(@foobar))
benjikins I didn't realize I could use any that way
thank you
timotimo m: my @pieces = <hello world how are you>; say "oh well this is weird".contains(any(@pieces))
camelia rakudo-moar 6a8278: OUTPUT«any(False, False, False, False, False)␤»
timotimo m: my @pieces = <hello world how are you>; say so "oh well this is weird".contains(any(@pieces))
camelia rakudo-moar 6a8278: OUTPUT«False␤»
timotimo m: my @pieces = <hello world how are you>; say so "oh well this is world".contains(any(@pieces)) 06:54
camelia rakudo-moar 6a8278: OUTPUT«True␤»
timotimo m: my @pieces = <hello world how are you>; say "oh well this is world".contains(any(@pieces))
camelia rakudo-moar 6a8278: OUTPUT«any(False, True, False, False, False)␤»
timotimo junctions are cool that way.
06:58 aindilis joined 06:59 labster left 07:07 leont joined 07:10 [particle]1 joined 07:11 sQuEE joined, aindilis` joined, peteretep_ joined 07:12 Ulti_ joined 07:13 Jonis_ joined 07:14 nemo_ joined, ranguard_ joined, rudi_s_ joined, bonsaikitten joined, zoosha_ joined, mr-fooba_ joined, luis` joined, El_Che_ joined, erdic_ joined 07:15 ribasushi_ joined 07:16 xiaomiao left, mr-foobar left, nemo left, ranguard left, Ulti left, faubi left, aindilis left, cgfbee left, erdic left, raydiak left, bitmap_ joined, erdic_ is now known as erdic 07:17 bitmap_ is now known as bitmap, bhm left, konobi left, shadowpaste joined 07:18 raydiak joined, peteretep_ is now known as peteretep, bhm joined, konobi joined 07:19 konobi is now known as Guest6604, markk joined, faubi joined, cgfbee joined 07:21 sergot joined 07:22 Ben_Goldberg left, RabidGravy joined 07:24 krunen joined 07:25 canopus left, smls joined
smls m: say "aaa {("bbb" for ^3)} ccc" 07:26
yoleaux 26 Aug 2016 21:58Z <tbrowder> smls: thanks! the "+$0" did the trick
camelia rakudo-moar 6a8278: OUTPUT«aaa bbb bbb bbb ccc␤»
smls m: say "aaa {"bbb" for ^3} ccc"
camelia rakudo-moar 6a8278: OUTPUT«Use of Nil in string context in block <unit> at <tmp> line 1␤aaa ccc␤»
smls ^^ Why does the for loop not return a value without parens there?
07:27 ufobat left
smls m: say "aaa {"bbb" if 1} ccc" 07:27
camelia rakudo-moar 6a8278: OUTPUT«aaa bbb ccc␤»
smls ^^ works for `if` but not `for`
TimToady loops at statementlist level are autosunk because people don't expect them to suddenly be lazy 07:30
yoleaux 25 Aug 2016 11:10Z <ab6tract> TimToady: do you by any chance have the notes for your ballistic programming talk(s) online?
smls TimToady: Ah, so {} in strings is statement-list context like a sub 07:31
TimToady m: say "aaa {do "bbb" for ^3} ccc"
camelia rakudo-moar 6a8278: OUTPUT«aaa bbb bbb bbb ccc␤»
07:31 canopus joined
TimToady do also works 07:32
smls oh, didn't know one could `do` a statement modifier
TimToady it's just a block, so there's a sequence of statements expected
do is a statement prefix
and modifier is part of a statement
it actually parses do ("bbb" for ^3) 07:33
07:39 leont left
smls TimToady: In the case of {} in strings, it wouldn't make the loop lazy though 07:46
07:47 labster joined
smls as it is immediately stringified 07:47
and I don't thing there's a use-case where you'd *want* to interpolate Nil into a string, as it just throws a warning 07:48
07:55 edehont joined 07:58 benjikins left 07:59 gfldex joined 08:00 espadrine joined 08:03 xtt joined 08:04 nadim left 08:05 vendethiel left 08:07 leont joined 08:08 xtt left
moritz \o 08:09
timotimo o/ moritz
i highlighted you instead of .tell; is that okay?
moritz timotimo: it's OK; an issue on github.com/moritz/ilbot/issues would have been even better :-) 08:11
timotimo ah, i didn't realize the ilbot repo would also be the right spot for the html pages
smls Does anyone know a "try Perl 6 online" thing that runs Perl 6.c?
ideone.com runs ancient parrot-based rakudo 08:12
timotimo did you see the pull request MasterDuke made to your ilbot repo?
moritz timotimo: the problem is that the search index I use goes through a stemmer, so a substring search on unstemmed words isn't easily possible
smls dev.perltuts.com/try runs 2015.11
timotimo there's the thing that perl6.party uses, but i forgot its name
moritz timotimo: I'd have to maintain a second search index
timotimo: yes, github.com/moritz/ilbot/issues/47 has the discussion 08:13
timotimo i didn't see it 08:14
MasterDuke: seems like a simple fix ought to be possible to make that pull request mergable
smls: glot.io/ - that one supports perl6 and has a new-ish version available 08:15
smls thanks
timotimo a 2016.04 moar rakudo
08:20 edehont left 08:38 RabidGravy left
Woodi open source vs lawyers approach in Linux from yesterday: lists.linuxfoundation.org/pipermai...03580.html 08:42
08:42 mvorg joined
Woodi looks open source is a not bit like traditional academia :) 08:43
08:48 domidumont left
Woodi btw. SSE4.2 have "Four string/text processing instructions providin 08:49
...g a rich set of primitives"
...for eg. XML parsing in few CPU instructions ? :) 08:50
08:52 leont left 09:03 ufobat joined 09:11 ufobat left 09:25 kjs_ joined 09:30 dataangel left 09:32 pecastro joined 09:58 telex left 10:00 telex joined, Ven joined 10:02 TEttinger left 10:29 rindolf left 10:33 pierre_ joined, rindolf joined 10:34 labster left 10:41 mcmillhj joined 10:46 pierre_ left, mcmillhj left 10:47 pierre_ joined 11:10 dalek left 11:11 dalek joined, ChanServ sets mode: +v dalek
timotimo MasterDuke: the coverity scans are only up to date because i just submitted a new scan :) 11:13
MasterDuke oh, ha 11:14
nice it was that easy 11:15
timotimo well, yeah. just had to ping [ptc] so he could add me to the project 11:17
11:18 kjs_ left
MasterDuke also, moritz++ cherry-picked the second commit in my ilbot PR. i'm still trying to find a better modification to the regex than my first commit that doesn't blow it up to an absurd size 11:18
11:20 espadrine left, ufobat joined
timotimo [ptc]: did you realize that "ptc" could also be short for "pretty cool"? 11:21
[ptc] timotimo: hehe :-) 11:22
yoleaux 06:17Z <timotimo> [ptc]: were you the one who got coverity for moarvm?
timotimo %) 11:23
11:23 mcmillhj joined 11:24 iH2O joined
MasterDuke timotimo: since evidence suggests you're awake now, this is me reminding you to re-run your coverage report 11:25
timotimo oh
11:28 mcmillhj left 11:30 Ven left 11:40 nemo_ is now known as nemo 11:41 iH2O left
domm what would I use to get something like Term::ReadKey ReadMode 2 in Perl 6 (i.e. a prompt without echo, eg for passwords)? 11:45
oh, and greetings from the train from Budapest to Vienna
mst domm: this is probably not the right answer, but due to ReadKeybeing XS, when writing fatpackable code I just do system('stty -echo'); ... system('stty echo'); 11:47
so you can always stick that in with a '# blame mst' and replace it with the right answer when somebody provides it
11:47 antipsychiatry joined
gfldex domm: see gist.github.com/gfldex/7d716436aa8...abd7106c3d 11:48
11:48 aries_liuxueyang left
gfldex mst: right answer ^^^ :) 11:48
mst \o/ 11:49
11:49 antipsychiatry left, kurahaupo joined, mcmillhj joined
gfldex domm: please note that '/dev/tty' may prove difficult on windows 11:49
timotimo gfldex: if you had a supply block there, you could have a phaser for when the supply gets closed, so that the "max" logic could wander into the whenever! :) 11:50
but the ENTER and LEAVE is nice, too
domm thx
mst wow but Term::termios has a bizarre calling convention
domm it's a private tool I use, so /dev/tty is no problem
gfldex mst: it surely does 11:51
it's a fairly thin layer on the c-stuff that shows its age
mst the C calling convention for getattr() and setattr() is not bizarre, just old-school C 11:52
I'm perfectly familiar with this
timotimo time to build Term::termios:api<2> :)
mst it's the way it's been translated that's making my eyes cross ;)
gfldex CLI::Promt::Password may actually what most ppl look for 11:53
11:54 mcmillhj left 11:55 smls left
domm I can try to implement CLI::Prompt::Password, but probably only next week (at Alpine Perl, with the help of Liz and nine) 11:55
11:57 kaare_ left
mst I'm guessing by the way gfldex said that that there is one somewhere? 11:57
gfldex there is github.com/titsuki/p6-Terminal-Readsecret what comes with a dep. to readsecret. No idea if that is portable. 11:59
mst ah, ok 12:01
domm good enough for me for now
12:03 Ven joined
domm another question: how can I convert a Buf (as returned by Digest::MD5.md5_buf) into an Int? 12:05
mst can you not .Hex.Int it or something? 12:07
domm reason: I calc a md5-sum (actually I'd prefer SHA1, but could not find it), convert the first few bytes to int, and seed the random number generator with that
12:07 cpage_ joined
domm mst: I tried, but was offline and did not get any results 12:07
But not sure if i tried .Hex.Int
mst I may bbe imagining it 12:08
domm Method 'Hex' not found for invocant of class 'Buf'
mst but it seems like you're trying to go buf -> hex -> int
MasterDuke the docs suggest 'use experimental :pack;' and then the unpack method 12:09
domm .Int returns the number of bytes in the buffer
MasterDuke: ok, thanks, I'll try that
(as I said I was offline and could only grep the source/docs for stuff, and google > grep)
nice, this works 12:12
and it seems that my offline docs are rather old..
MasterDuke the docs have seen a lot of action the last couple months: github.com/perl6/doc/graphs/commit-activity 12:19
12:26 jameslenz left, ufobat left 12:28 jameslenz joined
domm use experimental :pack; use Digest::MD5; my $d = Digest::MD5.new; $d.md5_buf(<"The","European","Perl","Conference">).unpack("CCCCCC").reduce: * ~ * 12:33
not exactly what I'm doing, but it works - thanks! 12:34
ah, and one last questions: how can do the folling Perl 5 code in 6 12:35
sub foo { say "foo" }; my $name = 'foo'; $foo->(); &$foo;
ah, soory 12:36
sub foo { say "foo" }; my $name = 'foo'; $name->(); &$name;
12:37 mcmillhj joined
domm perl -E 'sub foo { say shift }; my $name = "foo"; $name->("A"); &$name("B")' 12:37
dalek href="https://modules.perl6.org:">modules.perl6.org: 1d2db63 | (Tom Browder)++ | templates/root/index.html.ep:
correct grammar
 12:40
12:42 rudi_s_ is now known as rudi_s, mcmillhj left
dalek rl6-most-wanted: 8501b88 | (Tom Browder)++ | most-wanted/modules.md:
add new module WIP
 12:48
rl6-most-wanted: 559712a | (Tom Browder)++ | most-wanted/modules.md:
fix typo
 12:50
12:52 mra90 joined 12:53 mra90 left
MasterDuke m: class A { method foo { say "foo" } }; my $name = "foo"; A."$name"() 12:57
camelia rakudo-moar 6a8278: OUTPUT«foo␤»
MasterDuke domm: ^^^ for methods at least
13:01 wamba left
tbrowder ref modules: I've been trying to use Digest::xxHash, which uses naticecast, and see that somehow p6 translates uint64 into a bigint. how can we preserve the unsignedness in p6? 13:02
gfldex m: sub foo(){ 'oi!' }; my $foo = '&foo'; say ::($foo).(); 13:04
camelia rakudo-moar 6a8278: OUTPUT«oi!␤»
gfldex domm: ^^^
m: sub foo(){ 'oi!' }; my $foo = '&foo'; say ::($foo)();
camelia rakudo-moar 6a8278: OUTPUT«oi!␤»
MasterDuke m: sub foo(){ 'oi!' }; my $foo = 'foo'; say ::("&$foo")(); 13:06
camelia rakudo-moar 6a8278: OUTPUT«oi!␤»
MasterDuke tbrowder: there are a couple tickets (e.g., RT #124294, RT #127210, RT #127144) about uint types, it looks like they aren't quite working as people expect 13:12
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=124294
Link: rt.perl.org/rt3//Public/Bug/Displa...?id=127210
Link: rt.perl.org/rt3//Public/Bug/Displa...?id=127144
masak hi, #perl6 13:13
domm gfldex: thanks, again 13:22
pierre_ m: say "Bonjour!" 13:30
camelia rakudo-moar 6a8278: OUTPUT«Bonjour!␤»
pierre_ i have a small misunderstanding with defined 13:31
and Nil, or any
m: say Nil.defined
camelia rakudo-moar 6a8278: OUTPUT«False␤»
pierre_ Nil is not defined, which is expected
however, if i use Nil as a key in a hash, and then
let me show
13:31 MilkmanDan left 13:32 MilkmanDan joined, Ven left
pierre_ m: my %h = Nil => 'should not be defined'; for %h.kv -> $k, $v { say $k;say $k.defined; } 13:32
camelia rakudo-moar 6a8278: OUTPUT«Nil␤True␤»
pierre_ is that normal ?
gfldex pierre_: Nil is not _the_ undefined value. Using Nil by hand will almost always result in bugs.
see: docs.perl6.org/type/Nil and docs.perl6.org/language/typesystem...efinedness 13:33
pierre_ yes,, i do not plan to use Nil
it's for a test case
13:33 wphtech joined
jnthn I think the confusion here though is that some-identifier => blah will auto-quote the left hand side as a string 13:33
m: say (Nil => 'what').perl
camelia rakudo-moar 6a8278: OUTPUT«:Nil("what")␤»
pierre_ i have a function, that i want to test against several values
gfldex my %h = Nil => 'should not be defined'; dd %h; 13:34
jnthn m: say (Nil => 'what').key.WHAT
camelia rakudo-moar 6a8278: OUTPUT«(Str)␤»
gfldex m: my %h = Nil => 'should not be defined'; dd %h;
camelia rakudo-moar 6a8278: OUTPUT«Hash %h = {:Nil("should not be defined")}␤»
pierre_ oh, :)
make sense
gfldex m: my %h{Any} = Nil => 'should not be defined'; dd %h;
camelia rakudo-moar 6a8278: OUTPUT«Hash[Any,Any] %h = (my Any %{Any} = :Nil("should not be defined"))␤»
pierre_ so if i want to store
13:34 wphtech left
gfldex m: my %h{Any}; %h{Nil}= 'should not be defined'; dd %h; 13:35
camelia rakudo-moar 6a8278: OUTPUT«Hash[Any,Any] %h = (my Any %{Any} = (Nil) => "should not be defined")␤»
gfldex m: my %h{Any}; %h{Nil}= 'should not be defined'; %h.keys.say
camelia rakudo-moar 6a8278: OUTPUT«(Nil)␤»
gfldex m: my %h{Any}; %h{Nil}= 'should not be defined'; %h.keys.put
camelia rakudo-moar 6a8278: OUTPUT«Use of Nil in string context in block <unit> at <tmp> line 1␤␤»
pierre_ right
pair is auto quoting
just for the context, here is what i wanted to do 13:36
m: sub f( Any $obj) { return 'undef' unless $obj.defined; return 'true' if $obj == True; return 'false';} ; my %h = True => 'true', False => 'false', Nil => 'undef'; 13:38
camelia ( no output )
pierre_ and use %h to run my tests, by adding entries there 13:39
13:41 TheGrimFandango joined
TheGrimFandango hi 13:41
pierre_ i will then use an array and take element 2 by 2
m: my @a = Nil, 'should be undef'; for @a -> $k, $v { say $k.defined } 13:42
camelia rakudo-moar 6a8278: OUTPUT«False␤»
pierre_ Hi
gfldex if you want to hand Nil around like that you will have to use binding
pierre_ really, it's just in one test case 13:43
ok, i think the correct solution
i'll wirte a separate test for Nil :)
gfldex m: my Int $i is default(1); say $i; $i = Nil; say $i.defined;
camelia rakudo-moar 6a8278: OUTPUT«1␤True␤»
pierre_ and not use a hash
actually, in my test, i'm using Any, not Nil 13:44
but should be the same, no?
gfldex not at all
pierre_ just to check defineness
not to use it
gfldex you can use Mu if you need to differentiate between Any and "some undefined value" 13:45
13:45 bonsaikitten is now known as DrEeevil
pierre_ m: my $undef; my %h = $undef => 'not defined'; for %h.kv -> $k, $v { say $k.defined } 13:45
camelia rakudo-moar 6a8278: OUTPUT«Use of uninitialized value of type Any in string context.␤Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful. in block <unit> at <tmp> line 1␤True␤»
MasterDuke timotimo: btw, wasn't there some functionality missing in parse_coverage_report.p6 that i said i'd add? 13:46
13:52 araujo_ joined 13:55 araujo left
tbrowder .tell MasterDuke thanks! 13:55
yoleaux tbrowder: I'll pass your message to MasterDuke.
14:03 araujo__ joined 14:05 buggable joined 14:06 araujo_ left 14:07 skink joined 14:10 skids joined
tbrowder jnthn: any plans to work on uint64 issues in near future? 14:12
m: my $undef; say $undef.defined 14:24
camelia rakudo-moar 6a8278: OUTPUT«False␤»
tbrowder m: my $u = 1; say $u; $ 14:25
camelia rakudo-moar 6a8278: OUTPUT«WARNINGS for <tmp>:␤Useless use of unnamed $ variable in sink context (line 1)␤1␤»
tbrowder m: my $u = 1; say $u; $u = Nil; say $u 14:26
camelia rakudo-moar 6a8278: OUTPUT«1␤(Any)␤»
tbrowder m: my $u = 1; $u = Mu; say $u 14:28
camelia rakudo-moar 6a8278: OUTPUT«(Mu)␤»
tbrowder m: my $u = Mu; say $u.defined 14:29
camelia rakudo-moar 6a8278: OUTPUT«False␤»
tbrowder m: my $u = Any; say $u.defined 14:30
camelia rakudo-moar 6a8278: OUTPUT«False␤»
mr-fooba_ is there any guide to get started with perl6 on js ?
14:32 RabidGravy joined
tbrowder m: my $u = Nil; say $u.defined 14:33
camelia rakudo-moar 6a8278: OUTPUT«False␤»
tbrowder so it looks to me that one can set a previously defined var to Mu, Any, or Nil to undefine it, no? 14:34
14:34 espadrine joined
moritz not really 14:35
m: my Int $x = 42; $x = Any
camelia rakudo-moar 6a8278: OUTPUT«Type check failed in assignment to $x; expected Int but got Any (Any)␤ in block <unit> at <tmp> line 1␤␤»
moritz m: my Int $x = 42; $x = Nil; say $x.perl
camelia rakudo-moar 6a8278: OUTPUT«Int␤»
tbrowder m: my $u = 1; $u = Nil 14:37
camelia ( no output )
ugexe ...
m: say Nil ~~ Any 14:38
camelia rakudo-moar 6a8278: OUTPUT«True␤»
tbrowder m: my $u = 6; say $u.defined; $u = Nil; say $u.defined
camelia rakudo-moar 6a8278: OUTPUT«True␤False␤»
tbrowder it looks like, if one doesn't explicitly type a defined var, then assigning Nil to it undefines it 14:39
ugexe or stores the value which may or may not be undefined 14:40
tbrowder ?? 14:41
geekosaur I think they dropped a word. Nil stores the default value
m: my Int $v is default(6); $v = 10; say $v; $v = Nil; $v.say 14:42
camelia rakudo-moar 6a8278: OUTPUT«10␤6␤»
tbrowder i'm just interested in ensuring a previously defined var can be forced to become undefined
is there a best-practice way to do that? 14:43
geekosaur m: my Int $v is default(6); $v = 10; say $v; $v = Int; say $v.defined 14:44
camelia rakudo-moar 6a8278: OUTPUT«10␤False␤»
tbrowder m: my $u = 4; if $u say 'def'; $u = Nil; if !$u say 'undef or false' 14:46
camelia rakudo-moar 6a8278: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3my $u = 4; if $u7⏏5 say 'def'; $u = Nil; if !$u say 'undef ␤ expecting any of:␤ block or pointy block␤ infix␤ infix stopper␤»
ugexe my $x is default(1); say $x; $x = Nil:U; say $x.defined; say $x.WHAT
m: my $x is default(1); say $x; $x = Nil:U; say $x.defined; say $x.WHAT
camelia rakudo-moar 6a8278: OUTPUT«1␤False␤Nil␤»
ugexe gets around Nil setting default value of 1 which would be defined 14:47
14:48 mcmillhj joined
tbrowder i don't see a practical difference for my typical use case 14:49
let me try the last case again co
ugexe if you want to ignore `is default` and any 3rd party who may set that trait, then no there is no difference 14:51
tbrowder m: my $x = 't'; say $x; say 'true' if defined $x
camelia rakudo-moar 6a8278: OUTPUT«t␤true␤»
tbrowder okay, thanks, i guess i just don't understand meta programming!! 14:52
geekosaur um
are you confusing/conflating defined with existence?
(because I see no metaprogramming here) 14:53
tbrowder too deep for my simple needs
14:53 mcmillhj left 14:56 pecastro left 15:07 mcmillhj joined, pecastro joined
pierre_ m: Blob.new( 65535.base(16).comb(2).map( {:16($_)} ) ).say 15:10
camelia rakudo-moar 6a8278: OUTPUT«Blob:0x<ff ff>␤»
pierre_ any shortcut to go from 65535 to Blob 0x<ff ff>
my solution looks a bit too much :) 15:11
15:12 mcmillhj left 15:15 khw joined 15:16 zacts joined 15:17 mcmillhj joined
pierre_ Blob.new( 65535 ) 15:25
m: Blob.new( 65535 )
camelia ( no output )
pierre_ m: Blob.new( 65535 ).say
camelia rakudo-moar 6a8278: OUTPUT«Blob:0x<ff>␤»
ugexe m: say Blob.new(65535 xx 2) 15:30
camelia rakudo-moar 6a8278: OUTPUT«Blob:0x<ff ff>␤»
pierre_ and my solution is broken on top of that
m: 65539.base(16).comb(2)
camelia ( no output )
pierre_ m: 65539.base(16).comb(2).say
camelia rakudo-moar 6a8278: OUTPUT«(10 00 3)␤»
pierre_ hum, interesting, with xx
geekosaur that doesn't actually work 15:31
it's truncating the top byte of 65535 to get 255 and putting that in the Buf
pierre_ oh right :)
it's doing Blob.new( 65535, 65535) 15:32
15:32 rudolfochrist joined 15:34 mcmillhj left 15:35 cyphase left 15:39 cyphase joined
pierre_ m: pack 'n', 65535 15:41
camelia rakudo-moar 6a8278: OUTPUT«Use of the 'pack' function is experimental; please 'use experimental :pack'␤ in block <unit> at <tmp> line 1␤␤»
15:42 mcmillhj joined
pierre_ m: 'use experimental :pack; pack('n', 65535).say 15:42
camelia rakudo-moar 6a8278: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> 3'use experimental :pack; pack('7⏏5n', 65535).say␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement m…»
pierre_ m: use experimental :pack; pack( 'n', 65535 ).say 15:44
camelia rakudo-moar 6a8278: OUTPUT«Buf:0x<ff ff>␤»
pierre_ experimental, but that would work
15:45 wamba joined 15:46 pecastro left 15:47 mcmillhj left, ufobat joined
zacts hi perl6 nerds 15:54
grondilu pack is one of the first function I've delved deeply into while learning perl. I'm surprised it's experimental in Perl 6.